Sufficient condition for Blackhole formation in spherical gravitational collapse

SUFFICIENT CONDITION FOR BLACKHOLE FORMATION IN SPHERICAL
GRAVITATIONAL COLLAPSE

ROBERTO GIAMBO, FABIO GIANNONI AND GIULIO MAGLI

ABSTRACT. A sufficient condition for the validity of Cosmic Censorship in spherical
gravitational collapse is formulated and proved. The condition relies on an attractive math-
ematical property of the apparent horizon, which holds if "minimal" requirements of phys-
ical reasonableness are satisfied by the matter model.

1. Introduction

Predicting the final state (blackhole or naked singularity) of the gravitational collapse
of an initially regular matter distribution in General Relativity is still, to a large extent, an
open problem, even in the special case of spherical symmetry. In fact, Penrose's "Cosmic
Censorship" conjecture ^ stating that "every physically reasonable collapsing matter dis-
tribution forms a blackhole" seems to be overruled by several examples, in which what is
likely to be a physically reasonable distribution of matter forms a visible null singularity
(for instance in the dust and in the Vaydia case, see ^ and references therein) or even a
timelike singularity, as in the case of gravitating clusters of rotating particles 0, |[
It is, therefore, necessary to obtain a rigorous formulation of censorship in mathematical
terms, in order to be able to translate the conjecture in mathematically demonstrable as-
sertions. In this formulation, the physics of the problem is obviously expected to play a
fundamental role. For instance, the weak energy condition has naturally to be expected as
one of the hypotheses of any cosmic censorship theorem.

In the present letter, we report on an investigation carried out recently devoted to the
application of tecniques of non-linear o.d.e. to the censorship problem in spherical sym-
metry. The main result is a non-existence theorem for geodesies meeting the singularity in
the past, which gives a sufficient condition for a spherically symmetric non-static metric to
represent blackhole formation. The condition can be used as a test bed for probing models
of collapse without knowing the explicit solution of the Einstein field equations.

2. The analytical framework in brief

We consider a spherically symmetric collapsing object in full generality. The matter
model can thus be any model compatible with the weak energy condition. The general,
spherically symmetric, non-static line element in comoving coordinates t, r, 8, ip can be
written in terms of three functions v, A, R of r and t only as follows:

ds 2 = -e 2 "dt 2 + e 2A dr 2 + R 2 (d6 2 + sin 2 9dp 2 ) ,

A fundamental quantity is the mass function m(r, t) defined through

(2 1) 1 _2jn = dROR

R J dx* dx v

that is

(2.2) m(r, t ) = ^[ 1 ~ {R'f^ 2X + (i?)V 2l/

where a dash and a dot denote derivatives with respect to r and t respectively. Field equa-
tions Gq = 87tXq and G r r = 8ttT£ relate derivatives of m to the energy density e and the
radial stress p r of the material:

SUFFICIENT CONDITION FOR BLACKHOLE FORMATION IN SPHERICAL GRAVITATIONAL COLLAPSE 2

(2.3) rri = 4tt€R 2 R'

(2.4) m = -Airp r R 2 R

We consider only matter configurations admitting a regular center, and we suppose al-
ways that the collapse starts from regular initial data on a Cauchy surface (t = 0, say), so
that the singularities forming are a genuine outcome of the dynamics. If the solution is ini-

tially regular, (2.3) implies that the energy density - and therefore the spacetime - becomes
singular whenever R or R' vanish during the evolution. It follows, that singularities can
be of two different kinds: shell crossing, at which R' vanishes while R is non-zero, and
shell focusing at which R vanishes. The shell crossing singularities have been frequently
considered as "weak" although no proof of extensibility is as yet available in the literature.
In any case, in most physically interesting situations such singularities do not occur, so that
we shall concentrate attention here only on the shell focusing case. Therefore, we assume
that R' is initially non zero for non-zero r and remains positive up to the formation of the
focussing singularities.

The locus of the zeroes of the function R(r, t) defines the singularity curve t s (r) by
the relation R(r, t s (r)) = 0. Physically, t s (r) is the comoving time at which the shell of
matter labeled by r becomes singular. The singularity forming at r = 0, t = t s (0) is called
central as opposed to those occurring at r = tq > 0, t = t s (ro).

A singularity cannot be naked if it occurs after the formation of the apparent horizon.
The apparent horizon (t/i(r), say) is the boundary of the region of trapped surfaces and

it is therefore defined, due to equation (2.1), by R(r,th(r)) = 2m(r,th(r)). Since R

vanishes at a singularity, any naked singularity in spherical symmetry must be massless; at
a massless singularity the horizon and the singularity form simultaneously. Since regularity
of the center up to singularity formation requires m(0, t) = OVi < i s (0), the center is
always a candidate for nakedness. Other points ("non-central points") of the singularity

curve can be naked only in presence of radial tensions. In fact, due to eq. (2.4), the mass
is increasing in time in a collapsing (R < 0) situation, if the radial pressure is positive

To analyze the causal structure, observe that, if the singularity is visible to nearby ob-
servers, at least one outgoing null geodesic must exist, that meets the singularity in the
past. Such a geodesic will be a solution of

(2.5,
where

(2.6) v{r>t) , = lir L = e x- v

V 5oo

with initial datum t(ro) = t s (r^). For a problem of this kind, in which the initial point is
singular (the function (p is not defined at (vq, t s (ro))) no general results of existence/non
existence are known. As a consequence, in the literature, an approach has been developed
(see e.g. |Q|) which makes use of 1' Hospital theorem to identify the possible values of the
tangent of the geodesic curve at the singularity. This is the approach that allowed a full
understanding of the dust and of the Einstein cluster cases recalled above. However, to be
successful, this tecnique requires complete integration of the field equations, a result which
is far beyond our present understanding even in the simple case of the barotropic perfect
fluid. On the other side, what actually enters the problem of the causal structure is only the
function <p, not the whole solution of the field equations. Therefore, one can consider this

problem as an existence/non-existence problem for the non-linear o.d.e. (2.5), in which the
mathematical structure of the Einstein equations as well as the physics of the problem (like
e.g. formation of trapped surfaces, weak energy condition) play a fundamental role.

SUFFICIENT CONDITION FOR BLACKHOLE FORMATION IN SPHERICAL GRAVITATIONAL COLLAPSE 3

3. The non-existence theorem

From now on we will assume the following condition on p r , e (see Remark 3.3 below):

1

(3.1)

e > , p r > Max{— e,

8irR 2

}•

3.1. Theorem. // (jy|) holds and §f < in a neighborhood of (r , t s (r )), with r > 0,

the singularity forming at (r , t s (r )) is covered.

The theorem is based on a remarkable mathematical property of the apparent horizon,
which we address in the following

3.2. Lemma. If (3. 1) holds then there exists r* > ro such that the apparent horizon th(r)
is a subsolution of (2.5) for r € (ro, r*).

Proof. Differentiating the equation R(r, th(r)) = 2m(r, th(r)) with respect to r we get

dth
dr

where

2m'-i?'_ i?'
2m -i? i?

T :=

1 - 87rei? 2
1 + 8np r R 2

and the field equations ( 2.3 )-( 2.4) have been used. On the other end, from eq. (2.2) we get

ip(r, t)

1

e 2v R- 2

1 -

2m
~R

R!_
R

where a (crucial) minus sign is due to the fact that we consider a collapsing scenario (thus R
is strictly negative, at least near the singularity). At t = t/, the quantity in square brackets
is equal to one so that ip(r,th) = —R'/R. It follows dth/dr < ip(r,th) whenever the
quantity V is less than or equal to unity, that is

8vri? 2 (e+p r )

(3.2)

8irp r R 2 + 1

> 0.

□

3.3. Remark. If 8np r R 2 + 1 is negative, the radial pressure would diverge to minus infinity
at a singularity, a manifestly unphysical situation. Thus, we consider further only matter
models satisfying p r > —1/8ttR 2 . This is a very weak bound; obviously it must be
considered only if tensions are present, and at most it has to be extended to the whole of
the collapsing object (in this case it suffices to require p r > — l/8irR 2 (ri 3: t) where rt is
the boundary of the object). Once this is satisfied, the inequality (3.2) holds if e + p r > 0.

Proof of Theorem 3.1. Let t p (r) the solution of £'(?') = (p(r,t(r)) such that t p (ro)
t s (ro). By contradiction we suppose the existence of r\ > such that t P (r\) < th(r\) and
tp( r ) < th( r )> G [ro, r{\. We can suppose r\ < r*, where r* comes from Lemma 3_2
Since t p (r ) = th(r ), one has

(3.3) < t h { ri ) - tp(n) = [t h { ri ) - t p {r x )) - ((t h (r ) - tp(r ))) =

(t' h (0-W) n = (t' h (0-<p(CM0))n,

where £ € (r , ri). Using Lemma 3.2 it is t' h (^) < ip(^,th (£)), and hence

(3.4)

Combining (3.3) and ( |3.4| ) one gets a contradiction if 0) <

□

SUFFICIENT CONDITION FOR BLACKHOLE FORMATION IN SPHERICAL GRAVITATIONAL COLLAPSE 4

4. Discussion and conclusions

It is important to analyze the relationship of the condition (3.1) of the former lemma
with the weak energy condition (wee). The latter requires e > 0, e+p r > and e+p t > 0.
Here, there is no condition on the tangential pressure. If the radial pressure is positive, then

(3.1) coincides with the remaining inequalities of wee, and is therefore weaker. In presence
of tensions (p r < 0), it places a lower bound on the radial stress which again coincides
with that coming from wee unless e > 1/8ttR 2 , when the lower bound on p r must be
explicitly required.

The condition stated is only sufficient. As a test-bed we can use the dust case, in which
the exact solution is known in closed form and the nature of the singularities is known
in full details ^ (a discussion of dust collapse in terms of existence/non existence of
solutions for non-linear o.d.e. can be found in [[)]]).

The solution space for marginally bound collapse can be parameterized in terms of an
integer (n, say) giving the order of the first non-vanishing derivative of the initial density
profile at the center (since p r is zero, the center is the unique point that can be naked). One
easily finds that, for n = 1 and n = 2, dip/dt is positive near the center, while for n > 2
one has

(4.1) §(r,*h(r)) = -- (1 - /3„r"- 3 ) + 0(r"- 2 )

at r

where (3 n is a positive quantity proportional to minus the first non vanishing derivative of
the density at the center. Thus, the singularity is certainly covered for any n > 4 and for
n = 3 if /?3 < 1. Actually, however, we know that the singularity is naked if n equals
one, if n equals two, or if n equals three but [3% is greater than a positive value (3 C which
in turn is greater than one [|[] ( (3 C — (26 + 15\/3) /4). Therefore, there exists is a region
of the solutions space in which blackholes still form, while our (thereby only sufficient)
condition does not hold. Interestingly enough, this phenomenon occurs - at least in dust
spacetimes - near to the transition of the critical parameter. Physically, it reflects the fact
that the absence of apparent horizon formation prior to singularity does not necessarily
implies nakedness [|To|].

The results of the present paper can be applied to all the matter models that admit regular
initial data in spherical symmetry. For fluid sources, the space of all such solutions can be
parameterized in terms of three functions, namely the initial density, the initial velocity,

and the equation of state 1 1 1 1. A physically viable formulation of a Cosmic Censorship
theorem for fluid bodies would therefore rely in a classification of this solution space in
terms of the final outcome of the collapse. In this context the theorem proved here can be
used to characterize a "large" subset of this space, that contains only blackholes. Work in
this direction is in progress.

Analyzing in full generality what happens at the "boundary" of this subset, i.e. when a
transition from blackholes to naked singularities is expected to occur, looks, unfortunately,
still a quite far objective.

References

[1] R. Penrose, Nuovo Cimento 1 252 (1969).

[2] P. S. Joshi, Global aspects in gravitation and cosmology, (Clarendon press, Oxford, 1993).

[3] G. Magli, Class. Quantum Grav. 15 3215 (1998).

[4] T. Harada, H. Iguchi and K. Nakao, Phys. Rev. D 58 R041502 (1998).

[5] Kudoh, H. Harada, T. Iguchi, H. Phys. Rev. D (3) (2000) 104016,

[6] S. Jhingan and G. Magli, Phys. Rev. D61 (2000) 124006

[7] F. I. Cooperstock, S. Jhingan, P. S. Joshi and T. P. Singh, Class. Quantum Grav. 14 2195 (1997).
[8] T. P. Singh and P. S. Joshi, Class. Quantum Grav. 13 559 (1996).
[9] R. Giambo' and G. Magli, preprint.
[10] Jhingan, S. Joshi, P. S.; Singh, T. P. Class.. Quantum Grav. 13 (1996) 3057.

[11] Jhingan, S. and Magli, G, In Recent developments in General Relativity B. Casciaro, D. Fortunate), A.
Masiello, M. Francaviglia ed., Springer Verlag (Berlin) (2000).

sufficient condition for blackhole formation in spherical gravitational collapse 5

dlpartimento di matematica e flsica
Universitadi Camerino, Italy

E-mail address: giambo@campus.unicam.it , giannoni@campus.unicam.it

DlPARTIMENTO DI MATEMATICA,
POLITECNICODI MlLANO, ITALY
E-mail address: magli@mate . polimi . it